White House chief of staff meets Anthropic CEO on Mythos AI as Washington weighs risks and benefits
White House chief of staff Susie Wiles met Friday with Anthropic CEO Dario Amodei to discuss the artificial intelligence company’s new Mythos model, a system that has drawn federal interest for its potential implications for national security and the economy.
A White House official, speaking before the meeting on condition of anonymity, said the administration is engaging with advanced AI labs about their models and software security, and stressed that any technology considered for federal use would first undergo a technical evaluation period.
After the meeting, the White House described the discussion as productive and constructive, saying participants explored opportunities for collaboration and emphasized the goal of balancing innovation with safety. The talks followed months of friction between the Trump administration and Anthropic, which has positioned itself as safety-focused and has sought guardrails on AI development.
President Donald Trump moved to halt federal agencies from using Anthropic’s chatbot Claude amid a contract dispute with the Pentagon, declaring in a February social media post that the administration “will not do business with them again!” Asked Friday in Arizona whether Anthropic had a meeting at the White House, Trump said he had “no idea.” Defense Secretary Pete Hegseth also sought to designate Anthropic as a supply chain risk, an unprecedented step against a U.S.
company that Anthropic has challenged in two federal courts. The company has said it wanted assurances the Pentagon would not use its technology in fully autonomous weapons or for the surveillance of Americans, while Hegseth said the company must allow any uses the Pentagon deems lawful.
In March, U.S. District Judge Rita Lin issued a ruling that blocked enforcement of Trump’s social media directive ordering all federal agencies to stop using Anthropic products. Anthropic announced Mythos on April 7, calling it “strikingly capable” and restricting access to select customers because the system could surpass human cybersecurity experts in identifying and exploiting software vulnerabilities.
Some industry figures have questioned whether the claims overstate the model’s power, but even prominent skeptics have urged caution. David Sacks, a frequent Anthropic critic who served as the White House’s AI and crypto czar, said on the “All-In” podcast that people should “take this seriously,” arguing that more capable coding models are, by extension, better at finding bugs and chaining vulnerabilities into exploits.
The model has also drawn scrutiny abroad. The United Kingdom’s AI Security Institute said it evaluated Mythos and found it to be a step up over previous models, which were already improving rapidly. Anthropic said Amodei’s meeting included senior administration officials and focused on how the company and the U.S.
government could work together on shared priorities such as cybersecurity, maintaining America’s lead in AI, and AI safety. The company said it looks forward to continuing the discussions. The White House has said that before any adoption of new systems, the government will conduct its own technical assessments.